Autoencoder and Its Various Variants
By Jun-Hai Zhai, Sufang Zhang, Junfen Chen, Qiang He. IEEE International Conference on Systems, Man and Cybernetics, 2018
301 citations recorded by the source corpus.
The concept of autoencoder was originally proposed by LeCun in 1987, early works on autoencoder were used for dimensionality reduction or feature learning. Recently, with the popularity of deep learning research, autoencoder has been brought to the forefront of generative modeling. Many variants of autoencoder have been proposed by different researchers and have been successfully applied in many fields, such as computer vision, speech recognition and natural language processing. In this paper, we present a comprehensive survey on autoencoder and its various variants. Furthermore, we also present the lineage of the surveyed autoencoders. This paper can provide researchers engaged in related works with very valuable help.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing
View Paper: https://www.semanticscholar.org/paper/7e9e2006ed05fb92986ede10a62c4050afe3b351
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Enas Elgeldawi
- Thinagaran Perumal
- Norwati Mustapha
- Aboul Ella Otifey Hassanien
- Tomasz Bednarz
- Liming Zhu
- Raihani Mohamed
- Rowan T. Hughes
In Computer Vision and Pattern Recognition
- Thinagaran Perumal
- Norwati Mustapha
- Aboul Ella Otifey Hassanien
- Tomasz Bednarz
- Liming Zhu
- Raihani Mohamed
- Rowan T. Hughes
- Farhad Shiri
Citations in the corpus
- Membership Inference Attacks against Generative Models with Probabilistic Fluctuation (unknown): “VAEs has the similar structure as an autoencoder [ Zhai et al. , 2018 ] , which is composed of two modules: probabilistic encoder q ϕ ( z | x ) and decoder p θ ( x | z ) .”
- Deep learning in drug discovery: an integrative review and future challenges (2022, background): “Because of the backlog, there are fewer types of data produced, and the builtin data structures are used to their fullest extent (Zhai et al. 2018).”
- Generative Adversarial Networks–Enabled Human–Artificial Intelligence Collaborative Applications for Creative and Design Industries: A Systematic Review of Current Approaches and Trends (2021, background): “These approaches have produced excellent results and are widely studied (Zhai et al., 2018).”
- A Comprehensive Overview and Comparative Analysis on Deep Learning Models: CNN, RNN, LSTM, GRU (2023, background): “stochastic functions, represented as pencoder (z|x) and pdencoder (r|z), respectively, where r denotes the reconstruction of x [53].”